Estimating the minimum sample size for interventional and observational studies using the lung clearance index as an endpoint✰

Abstract

Background: With the increasing availability of highly effective modulators for people living with cystic fibrosis (CF), there is a need to re-design research studies to reflect the changing epidemiology of the CF population. The lung clearance index (LCI), a sensitive physiological measure of lung function, may be ideally suited as an endpoint in the era of CF modulator therapies. In this study we describe study design considerations for implementing LCI into interventional and observational research.

Methods: Simulations were used to estimate the required sample size to detect a range of treatment effects for interventional studies (including cross-over trials) and to track lung disease progression in observational studies.

Results: Using published treatment effects to inform the design of prospective studies can lead to inefficient study designs. Large improvements in LCI for a few individuals can skew results and can influence interpretations of treatment effects. Adjusting for baseline LCI can help to improve the efficiency of a study. Compared to the forced expiratory volume in 1 second (FEV₁), analysis using LCI as an endpoint requires as little as one third of the total sample size.

Conclusions: Planning of prospective studies that include LCI as an endpoint need to consider baseline LCI and disease severity of the study population; whereas interpretation of results needs to consider whether a few individuals skew the overall treatment effect.

Keywords: Cystic Fibrosis; Lung clearance index; Sample size; Study design.